Advanced Stirling Converter Heater Head Creep Testing by dandanhuanghuang


									             Ethan Brewer
      The Ohio State University
Aeronautical / Astronautical Engineering
       Undergraduate, 2nd Year
Mentors: Dave Krause and Ralph Pawlik
 Background on Stirling Engine
 Testing Concerns
 Creep Testing Setup
 Problems in Testing
 Solutions to Problems
 Data Analysis
 Other Projects
 Conclusion
    The Advanced Stirling
Radioisotope Generator (ASRG)
   Idea of Stirling Engine goes back several centuries
   ASRG converts heat energy from a General
    Purpose Heat Souce (GPHS) to mechanical and
    then electrical energy
   Uses two Advanced Stirling Converters (ASCs)
    placed end to end
   Intended primarily for use on deep space missions
    to power space vehicles; may also be used in
    lunar or Mars missions
   To be implemented on a mission as early as 2014
          Background (cont.)
 Project being developed
  jointly under Department
  of Energy, Lockheed
  Martin, NASA, and
  Sunpower, Inc.
 ASRG more efficient than
  current Stirling engines /
  other radioisotope
  thermoelectric generators
 Must last 17+ years with
  no maintenance
     The ASC

Heater Head
Hot End
Stirling Working Space
(thick wall)

                   Cold End
                   (thin wall)
ASC in Action
Heater Head C08
        Primary Concerns
 Heater head conducts heat into ASC,
  hot end reaches high temperature (over
  800 °C)
 Also highly pressurized with Helium
 Made of super-alloy to withstand the
  high temperature, pressure, and
 Must endure conditions for lifespan
            Creep Testing
 Specialized rigs designed to heat and
  pressurize heater heads to test
  temperature and pressure around thin-
  walled area (accelerates creep)
 Laser micrometers positioned on X and Y
  axes to measure creep strain in-test (plus
  for pre- and post-test complete scans)
 Thermocouples (T’cpls) welded to heater
  head to set and monitor temperature
 Intended to operate for about a year in
  accelerated creep testing
Test Setup
   Test article SN18
    creep test ended
    prematurely (only a
    few days testing)
   Microcracking
    around welded
    propagated to cause
   Need new
    temperature monitor
    and control method
   Undertook study to
    test contact T’cpls
    (Type R)
   Developed method
    using contact T’cpls
    plus welded T’cpls on
    unpressurized crown
    to monitor and control
   Readings more
    accurate than type K’s
    with help of statistical
             Solution (cont.)
   Basic statistics
    based on normal
    distribution (right)
   Formulas based on
    this curve and
    standard deviation
    give Standard Error
    of the Mean
   Allow to take a small
    sample set with
    confidence in results
It Works, and Looks Good
 Before           After

   Mooney 2-10            C08

          Analyzing Data
 Problem with how to analyze the long
  sets of data gathered (several hundred
  lines each)
 Data sets contain LVDT readings and
  Laser diameter readings
 LVDT readings do not line up from scan
  to scan
 How to find elastic and creep strains at
  varying axial positions?
       Analyzing Data (cont.)
   Previously used
    polynomial fit curves                         N-S Laser Strain Rate
    of several degrees                            E-W Laser Strain Rate
                                                  Analytical Strain Rate

                            Axial Position, mm
   Produced
    acceptable but
    idealized curves; no
    noise, smoothed
    features                                                     axial creep deformation removed

                                                 Creep Strain Rate, mm/mm x10 -6/day
   More accurate
    method desired
        Analyzing Data (cont.)
   Created and
    developed MATLAB
    script to better
    analyze data
   Uses interpolation to
    find pre-test and post-
    test diameters at
    uniform axial positions
   Much more realistic
    curves produced; can
    see features like noise
    (important for
    probabilistic analysis
    to be performed by
               Other Projects
                                  Planar Spring
   Planar Spring
    Fatigue testing for
   University of Akron
    will be performing
    tests using high
    frequency shakers
   Spring material
    samples shaken
    until they break, life
                             (example – not actual design)
        Other Projects (cont.)
   ARES I-X load link
    testing on Instron
    50-ton frame
   Tested deformation
    of link in tension and
    compression (up to
    50 kips)
   Controller software
    bug found,
    investigated, and
 Using new temperature measurment
  method and MATLAB creep data
  analysis script more accurate data can
  be gathered and analyzed in shorter
  time frame
 Better and more timely life prediction
 Adaptable for future testing
Special Thanks To:
 LERCIP Program Staff
 Dave Krause
 Ralph Pawlik
 Frank Bremenour
 All of you for attending

To top